Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 www.elsevier.com/locate/jembe

Chemical and physical defenses against predators in Cystodytes (Ascidiacea)

Susanna Lo´pez-Legentil a,b,*, Xavier Turon a, Peter Schupp c

a Department of Biology (Invertebrates), Faculty of Biology, University of Barcelona, 645 Diagonal Ave., 08028 Barcelona, Spain b Laboratoire de Chimie des Biomole´cules et de l’Environnement (Centre de Phytopharmacie), University of Perpignan, 52 Paul Alduy Ave., 66860 Perpignan, France c Marine Laboratory, University of Guam, UOG Station, Mangilao, GU 96923, Guam, USA Received 4 July 2005; received in revised form 1 November 2005; accepted 1 November 2005

Abstract

Ascidians utilize both physical (spicules, tunic toughness) and chemical defenses (secondary metabolites, acidity) and suffer relatively little by generalist predators. The genus Cystodytes (Polycitoridae) is distributed widely in both tropical and temperate waters. Secondary metabolite composition, calcareous spicules and tunic acidity (pHb1) may act as redundant defense mechanisms against predation in this genus. To assess the relative importance of chemical and physical defenses against predation in ascidians, we studied purple and blue morphs of Cystodytes from the western Mediterranean (formerly assigned to Cystodytes dellechiajei, but recently shown to belong to two different species), and a purple morph from Guam (USA), identified as Cystodytes violatinctus. Crude extracts, spicules, ascididemin (the major alkaloid of the blue morph) and acidity were used in feeding trials to evaluate chemical and physical defense mechanisms in Cystodytes spp. We performed feeding experiments in the field with a guild of generalist fish (mostly damselfish), and in the laboratory with a and a puffer fish. Our results showed that all crude extracts and ascididemin significantly deterred fish predation, but not sea urchin predation. However, neither acidity alone nor spicules at natural concentrations deterred feeding. These results and other studies on sponges and gorgonians suggest that secondary metabolites are the primary means of defense against fish predators. Spicules and tunic acidity may perform other ecological roles and/or target certain specialist predators. D 2005 Elsevier B.V. All rights reserved.

Keywords: Acidity; Ascidians; Ascididemin; Cystodytes; Defense mechanisms; Deterrence; Predation; Toxicity

1. Introduction range of defense mechanisms, including behavioral (e.g. mimicry), physical (e.g. spicules, tissue toughness) Benthic marine invertebrates are under intense com- and chemical (e.g. bioactive secondary metabolites, petitive pressure for space, light and nutrients. In addi- acid pH) strategies, to ensure survival. The relative tion, sessile organisms are easy prey for predators. importance and potential interactions of different de- Many of these organisms, therefore, have developed a fense mechanisms may depend on: the organism exam- ined (Pennings and Paul, 1992; Hay et al., 1994), the inability of a single defense to deter all type of pre- * Corresponding author. Present address: Biological Sciences and Center for Marine Science, University of North Carolina at Wilming- dators and/or competitors (Paul and Hay, 1986; Hay et ton, 5600 Marvin K. Moss Lane, Wilmington, NC 28409, USA. al., 1987; Schupp and Paul, 1994; Pisut and Pawlik, E-mail address: [email protected] (S. Lo´pez-Legentil). 2002; Tarjuelo et al., 2002; Burns et al., 2003), the

0022-0981/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jembe.2005.11.002 28 S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 organism’s life history stage (Uriz et al., 1996; Pisut predation (Parry, 1984; Tarjuelo et al., 2002). All in all, and Pawlik, 2002), the developmental or physiological little experimental evidence exists regarding the inter- constraints that might prevent a particular defense from action of chemical and physical defense mechanisms in being used in all parts of the organism, necessitating ascidians. additional defenses to ensure protection of the whole We chose the widely distributed ascidian genus organism (Harvell and Fenical, 1989), evolutionarily Cystodytes (Aplousobranchiata: Polycitoridae) to assess established constraints (Schmitt et al., 1995; Kubanek the relative importance of physical and chemical et al., 2002) and whether structures or compounds are defenses against predation in ascidians. Lo´pez-Legentil energetically expensive, among other factors. For in- et al. (2005a) identified two chemotypes within Medi- stance, in some sponges, both large spicules and chem- terranean specimens of this genus, previously attributed ical defenses deter fish feeding (Pawlik et al., 1995; to Cystodytes dellechiajei (Della Valle, 1877). Genetic O’Neal and Pawlik, 2002; Burns et al., 2003; Burns and and biological evidence indicates that these two che- Ilan, 2003). Soft corals and gorgonians produce both motypes correspond to sibling species (Lo´pez-Legentil chemical defenses and sclerites, which, in some cases, and Turon, 2005; Lo´pez-Legentil et al., 2005b). The protect them against predators (Pawlik et al., 1987; first chemotype was characterized by the presence of Harvell et al., 1988; Wylie and Paul, 1989; Pawlik C9-unsubstituted pyridoacridines such as ascididemin and Fenical, 1992; Puglisi et al., 2000, 2002; O’Neal (Fig. 1; Kobayashi et al., 1988) and 11-hydroxyascidi- and Pawlik, 2002). However, the primary function of demin (Schmitz et al., 1991), which were present in sclerites may be structural rather than defensive (Koehl, blue and green morphs from the western Mediterra- 1982; Lewis and Von Wallis, 1991). In view of these nean. The second chemotype contained the sulfur-con- and similar results, Van Alstyne et al. (1994) suggested taining pyridoacridines kuanoniamine D (Caroll and that, when feeding experiments are designed, all de- Scheuer, 1990), shermilamine B (Carroll et al., 1989) fense mechanisms against predation that an organism and their deacetylated forms (Eder et al., 1998; and might use should be tested in combination, rather than Lo´pez-Legentil et al., 2005a). These pyridoacridines each single defense separately. However, mainly due to were present in a purple morph from the western Med- experimental limitations, few studies have attempted iterranean, and were also found in a purple morph from this (Paul and Van Alstyne, 1992; Schupp and Paul, Guam, identified as Cystodytes violatinctus Monniot, 1994; O’Neal and Pawlik, 2002; Hill et al., 2005). 1988 (Lo´pez-Legentil, 2005). Some of these substances Among benthic invertebrates, ascidians suffer rela- are highly cytotoxic (Eder et al., 1998; Dassonneville et tively little predation by generalist predators (Millar, al., 2000; Bowden, 2000). Ascididemin also has some 1971; Goodbody and Gibson, 1974; Stoecker, 1980a). antibacterial and antifungal action (Lindsay et al., These generalist predators are mainly fish (Randall and 1995). To our knowledge, apart from a possible anti- Hartman, 1968; Myers, 1983), and occasionally urchins fouling function (Debard et al., 1998), no ecological (Briscoe and Sebens, 1988). Specialist predators on role has been described for ascididemin. In addition to ascidians include mollusks, such as the lamellarians, the bioactive compounds cited above, Cystodytes spp. cypraeids (Fretter and Graham, 1962; Millar, 1971; colonies are highly acidic (pHb2; Parry, 1984; Tarjuelo Lambert, 1980) and nudibranchs (Millar, 1971; Paul et al., 2002) and contain calcareous spicules encasing et al., 1990), and polyclad (Millar, 1971; the zooids (Turon, 1987; Kott, 1990). These potential Morris et al., 1980; Schupp et al., 1999; Caralt et al., defenses seem to result in an effective defense mecha- 2002). In addition, ascidians have both physical (spicules, 2 1 tunic toughness) and chemical defenses (Swinehart et 3 al., 1974; Stoecker, 1978, 1980b; Pisut and Pawlik, N 2002; Tarjuelo et al., 2002). Numerous studies support 4 N the deterrent role of secondary metabolites (Paul et al., 5 1990; Davis, 1991; McClintock et al., 1991; Lindquist 6 et al., 1992; Vervoort et al., 1998). High vanadium 9 8 concentrations (Stoecker, 1980b) and low pH (Webb, N 7 1939; Stoecker, 1978, 1980a,b,c; Pisut and Pawlik, 2002) have also been suggested as anti-predatory O defenses. However, other studies indicated that neither Fig. 1. Chemical structure of ascididemin, the major alkaloid of the the presence of vanadium nor an acidic pH prevented Mediterranean blue morph. S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 29 nism (although their relative importance and interac- were weighed and freeze-dried. They were then tions are not known), because Cystodytes spp. colonies extracted three times in a 1:1 (v/v) mixture of dichlor- generally lack epibionts and show scarce signs of pre- omethane and methanol and the combined extracts were dation. As the vanadium concentration in Cystodytes concentrated under vacuum to leave a powdery organic (V10 ppm dry weight; Stoecker, 1980b) is well below residue. The mean natural concentration of crude ex- the concentration suggested to deter predation in other tract found was 2.01% (F0.21 S.E.M., N =5 for each ascidians (N1000 ppm dry weight; Stoecker, 1980b), it color morph) per wet mass of the ascidian. We used this will be ignored in the present study. Tarjuelo et al. concentration, relative to hydrated artificial food, in the (2002) observed a deterrent activity of the crude extract assays. We also tested ascididemin (the major alkaloid of a blue Mediterranean morph against the hermit crab of the blue morph) synthesized in the University of Cestopagurus timidus. Perpignan, following Bracher (1989). The ascididemin To assess the relative importance of chemical and concentration used in feeding assays was 0.019% wet physical defenses against predation in ascidians, we weight, which was in the lower half of concentrations studied a purple and a blue morph from the western found by Lo´pez-Legentil (2005) after monitoring a Mediterranean, corresponding to the two chemotypes natural population over 2 years, so our results would explained above, and a purple morph assigned to C. be, if any, more conservative. violatinctus (Lambert, 2003) from Guam (USA). Crude Spicular weight per wet mass was calculated by extracts, ascididemin (the major alkaloid of the blue averaging the concentration found in 4 independent morph), spicules, and tunic acidity, as well as combina- colonies per morphotype. A piece of each colony was tions of these, were tested against a naturally occurring weighed, freeze-dried and then boiled in commercial guild of generalist fish (mainly the damselfish, Abudef- bleach for several minutes until the tissue dissolved duf vaigiensis and A. sexfasciatus) in the field, and the completely. Then, spicules were washed several times urchin savignyi and the puffer fish Canthiga- in water to remove organic remains, dehydrated in ster solandri, both in the laboratory. absolute ethanol and finally dried in an oven at 60 8C. The average spicular concentration found and 2. Materials and methods used in our assays was 0.64% (F0.01 S.E.M.) of the wet mass of the ascidian. Several colonies of each 2.1. Ascidian samples morphotype were processed to obtain enough spicules to conduct the experiments. Specimens of the purple and blue morph of the It has been recognized that performing anti-predato- ascidian Cystodytes were sampled by scuba diving in ry assays using dry mass ratios to determine concentra- Catalonia (NE Spain, western Mediterranean) in 2002 tions of metabolites (or spicules) may be misleading, as and 2003. For a morphological description, see Lo´pez- predators consume hydrated material (Harvell et al., Legentil and Turon (2005). The purple C. violatinctus 1988; O’Neal and Pawlik, 2002) and differences in was collected in Togcha Channel on Guam (NW Pacif- dry weight/wet weight ratios may translate into impor- ic, USA) in 2003 and identified according to Monniot tant differences in the effective concentrations assayed (1988) and Lambert (2003). All forms had disc-shaped when using dry weights. Volumetric concentrations and calcareous spicules, ca. 1 mm in diameter. The purple wet weight concentrations do not suffer from this prob- Mediterranean morph had, in addition to discoidal spi- lem. The two coincide when the density of the source cules, small spherical spicules ca. 70 Am in diameter organism and the artificial food is the same. We chose (Lo´pez-Legentil and Turon, 2005). The spicules added to work with wet weight concentrations because, as- to the artificial food were the discoidal ones or the suming that a predator consumes a given weight of discoidal plus the spherical ones, as indicated in each hydrated material per day, wet weight ratios may accu- test. rately reflect the amount of metabolites/spicules that a predator ingests. We performed comparative assays 2.2. Chemical extraction, spicule separation and with ascididemin using volumetric and wet weight concentrations assayed ratios and found similar outcomes (see Results). In any case, as the densities of our artificial foods were Biologically active compounds of Cystodytes are slightly lower (cubes for field assay=1.06 g mlÀ 1; sea restricted to the least polar fraction (Becerro et al., urchin food=1.16 g mlÀ 1; puffer fish food=1.15 g 1997), and are mainly pyridoacridine alkaloids (Bon- mlÀ 1) than that of the ascidian itself (1.23 g mlÀ 1), temps, 1996; Lo´pez-Legentil et al., 2005a). Colonies our results are conservative as we would be adding 30 S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36

Table 1 ropes. Each rope contained either 4 controls or 4 treated Artificial food recipes for damselfish, puffer fish and sea urchins food cubes. We placed a total of 24 pairs of control and Recipes Damselfish Puffer fish Urchins treated ropes on the reef, one pair at a time. In all Agar (g) 1.25 0.15 0.25 treatments, each pair was removed when approximately Carageenan (g) 1.25 0.15 0.25 half the cubes had been eaten. We used Wilcoxon’s Catfish (g) 5 2 2 signed-rank test for paired comparisons to test for Enteromorpha sp. – – 1 Water (ml) 80 18 25 significant differences in the number of cubes eaten between control and treated food. For the acid treat- Agar (Sigma no. A-6924), carageenan (Sigma no. C-1013), Kruse’s Perfection Brand catfish and freeze-dried Enteromorpha sp. ments, the whole experiment was limited to 35 min and a mean of 20 pairs of ropes because the acidity of the cubes changed from pH 1.5 at time 0 to pH 3.5 after 40 16%, 6%, and 7% more metabolites/spicules, respec- min in seawater (Fig. 2). To measure pH of the cubes, tively, had we worked with volumetric ratios. we followed the method of Davis and Wright (1989). Each cube measured was placed on SigmaR pH test 2.3. Field assays – generalist predators strips (precision=0.5 units) and pressure was applied until fluids were extruded. We then monitored the Field bioassays were conducted at a depth of 8 m at resulting color change of the paper. When acid and Gun Beach (Guam, USA). We conducted a feeding spicules were combined, a slight bubbling in the deterrence experiment using the field method of Schupp cubes was observed under the stereomicroscope. How- and Paul (1994). The following components were ever, although some calcium carbonate was probably added to an artificial diet at the wet weight concentra- dissolved with the acid, no change in the spicular shape tions mentioned above (see Table 1): crude extract, was observed and the pH increased only slightly at time ascididemin, disc-shaped spicules alone or with spher- 0 (~0.5). The addition of crude extract or ascididemin to ical spicules, and combinations thereof. For compari- the artificial food did not have any obvious effect on the son, the test with ascididemin was run at volumetric and original color of the food. wet weight concentrations. In addition, to test the role of low pH, 800 Al of sulfuric acid were added to obtain 2.4. Laboratory feeding assays – benthic predators pHb1. Preliminary trials were run to monitor the changes in acidity over time in acid-treated food. The sea urchin D. savignyi is very common on coral When solvents were required to dissolve crude extracts reefs around Guam and is normally found in shallow or ascididemin in the treated food, the same amount of waters (0.5 m to 3 m). Sea urchins are mainly grazers solvent was added to the control. The mixture was that feed on algae, but are usually referred to as gener- poured into 1-cm3 molds containing a rubber O-ring, alist predators that also feed on a variety of invertebrates which allowed the use of safety pins to attach cubes to (Barnes, 1987; Briscoe and Sebens, 1988). D. savignyi

7 Artificial food damselfish 6 Artificial food puffer fish

5

4 pH 3

2

1

0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 Minutes

Fig. 2. pH variation of artificial food over time. Vertical straight line indicates the ending time of the in situ acid tests with damselfish. Vertical dashed line shows the end time of puffer fish tests. Bars indicate standard errors. S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 31 adapts poorly to laboratory conditions and has to be kept squares of the window screening. Puffer fish were in separate tanks to avoid mortality (up to 75%, if they placed separately in 30-l flow-through tanks and nor- are kept together). Therefore, each sea urchin was mally started feeding on control food within one to two placed in a 30-l flow-through tank and repeatedly of- weeks. The tanks were cleaned regularly to prevent fered artificial food until it was used to feeding on it. The algal growth. Prior to the experiment, the fish were tanks were regularly cleaned to prevent algal growth. starved for 48 h to homogenize their state. Experimen- Artificial food was prepared, as indicated in Table 1. tal sets were composed of 14 fish, and were offered Treatments were prepared by addition of one of the either treatment(s) or control screens (7 replicates each). following: crude extract, ascididemin and disc-shaped The fish were allowed to feed for 1 h, in which time spicules, at the indicated wet weight concentrations. The approximately half the squares were eaten in any of the mixture was heated and poured into a mold backed with treatments. If the treatments contained sulfuric acid, the aluminum window screening to form a strip that covered fish were allowed to eat for 30 min, as pH reached a mean of 325 squares of the window screening. Prior to values of ~ 3 after that time (the pH of 2 extra screens experiments, the urchins were not fed for 24 h. Although submerged in empty tanks was recorded at the end of it is known that some change their feeding each test; Fig. 2). We used either t-tests or Mann– preferences when they are starved, we used this fasting Whitney analysis (if the assumptions of normality or period to homogenize the state of the sea urchins. A heteroscedasticity failed) to identify significant differ- deterrent food (note that we are performing deterrence ences between the number of squares eaten in the assays, not preference ones in which test animals have control and in the treatment. food choice) should not be consumed, even if the test organism is starved. If there were any discrepancy, the 3. Results results using slightly starved urchins would be more conservative. Experimental sets consisted of 14 urchins 3.1. Field assays – generalist predators that were offered either treatment or control screens (7 replicates each). Urchins were allowed to feed for one At Gun Beach (Guam, USA) the damselfish Abu- day and two nights. We used either t-tests or Mann– defduf vaigiensis and Abudefduf sexfasciatus were the Whitney analyses (if the assumptions of normality or main consumers during experiments, although there heteroscedasticity failed) to identify significant differ- was occasional consumption by the wrasse Thalassoma ences in the number of squares eaten between control lutescens and the triggerfish Balistapus undulatus. and treated food. When a new treatment set was placed, fish nibbled Due to the duration of this experiment, assays with randomly the cubes presented to them; therefore, the acid were precluded. Prior to experiments, we per- final choice did not depend upon which cube was formed some preliminary tests to ensure (via TLC) tasted first. Ascididemin, which was tested in both that crude extract and ascididemin were retained in volume and wet mass ratios, and ascididemin com- the artificial food for the entire experiment. bined with spicules deterred feeding by reef damselfish The puffer fish Canthigaster solandri is a benthic ( p b0.001 in all cases; Fig. 3). The consumption of predator that feeds on benthic algae, ascidians, and treated vs. control food in ascididemin treatments at other invertebrates (Amesbury and Myers, 1982). It volume and wet weight ratios was virtually the same adapts well to laboratory conditions and has been pre- (0.60F0.07 and 0.59F0.10, respectively; mean- viously used for deterrence tests (Rogers and Paul, FS.E.M.). The three crude extracts deterred fish feed- 1991; Becerro et al., 1998). Artificial food was pre- ing significantly more than controls ( p b0.001 in all pared as indicated in Table 1. Treatments were prepared cases; Fig. 3). The disc-shaped spicules, which are by addition of crude extract, ascididemin, disc-shaped typical of the genus, alone or combined with the spicules and some combinations thereof (each at wet spherical spicules found in the purple Mediterranean weight concentrations). For treatments involving acid- morph, did not deter predation ( p =0.48 and p =0.56, ification, 150 Al sulfuric acid was added to obtain respectively). Cubes with sulfuric acid or a combina- pHb1. The changes in acidity over time in acid-treated tion of sulfuric acid with spicules did not deter fish food were monitored in preliminary assays. The mix- feeding. In addition, when we added sulfuric acid to ture was heated for 30 s in a microwave, and after ascididemin, or to ascididemin and spicules, no deter- addition of the defense mechanism tested, the whole rent effect was found ( p =0.25 and p =0.07, respec- was poured into a mold backed with fiberglass window tively). However, when cubes were treated with screening to form a strip that covered a mean of 277 sulfuric acid, the fish displayed flushing behavior in 32 S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36

5 Control * * Treatment 4 * * * *

3

2

Number of cubes eaten 1

0

S SA G S Guam Sp G + Sp G CE CE MediCE Blue Sp G + Asc + SA Asc VolumeAsc Weight Asc + Sp G SA + SpSA + Sp Asc + SA

Fig. 3. Number of cubes eaten for each treatment by the damselfish Abudefduf vaigiensis and Abudefduf sexfasciatus during the non-choice assays. Codes for treatments: CE Guam, crude extract of the purple Guam morph; CE Medi, crude extract of the purple Mediterranean morph; CE Blue, crude extract of the blue Mediterranean morph; Asc Volume, ascididemin in a volume ratio; Asc Weight, ascididemin in a mass ratio; Sp G, disc- shaped spicules typical of the genus; Sp S, spherical spicules found in addition to disc-shaped ones in the purple Mediterranean morph; SA, acidified with sulfuric acid. Asterisks show significant differences. Bars indicate standard errors. which they repeatedly swallowed and regurgitated the All crude extracts tested, whether alone or in com- food cubes. bination with acid or spicules, significantly deterred predation by the puffer fish Canthigaster solandri 3.2. Laboratory feeding assays – benthic predators (Fig. 5). As all fish sampled the offered food, in no case was any negative result (zero consumption) D. savignyi showed no significant differences in recorded. Ascididemin, alone or with spicules, also consumption of control food or in any of the treatments significantly deterred predation ( p =0.023 in both offered (Fig. 4). Indeed, the urchins appeared to move cases), although, as happened during field experiments randomly in their tanks at night and eat whenever they with the damselfish, the addition of sulfuric acid found a screen, regardless of its content: aluminum seemed to counteract somehow the deterrent effect of screens were sometimes damaged by urchin bites. A the alkaloid plus spicules, yielding no significant deter- period of two nights was needed for at least 90% of the rent effect ( p =0.073). No significant differences were sea urchins to encounter the screen. found between consumption of control and treated food

1,2 Control Treatment 1,0

0,8

0,6

0,4

0,2 Squares eaten / Total squares

0,0

edi Asc uam Blue Sp G E M CE CE G C

Fig. 4. Number of squares eaten by the sea urchin Diadema savignyi per total number of squares during the non-choice assays. X-axis labels as in Fig. 3. Bars indicate standard errors. S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 33

0,7 Control * 0,6 Treatment * * * 0,5 * * * 0,4 ***

0,3

0,2

Squares eaten / Total squares 0,1

0,0

Asc Sp SA + SA CE GuamCE MediCE Blue SA + Sp Asc + Sp CE BlueCE +Medi SACE +Blue Sp + Sp CE GuamCE Medi + SA Asc + SA + Sp

Fig. 5. Number of squares eaten by the puffer fish Canthigaster solandri per total number of squares during the non-choice assays. Sp indicates disc-shaped spicules characteristic of the genus; the other X-axis labels are as in Fig. 3. Asterisks show significant differences. Bars indicate standard errors. containing spicules, acid or a combination of the two the mixture palatable. This loss of deterrence might be (pN0.1 in all cases). due to the nitrogen atoms of ascididemin becoming protonated at low pH. Indeed, in the living ascidian, 4. Discussion acidic vacuoles are located in bladder cells (Webb, 1939; Hirose, 1992), metabolites seem to be stored in Our study showed that crude extracts of the purple pigment cells (Turon et al., 2005), and spicules are and blue Mediterranean morphs of Cystodytes, and of found in the tunic matrix encasing the zooids (Turon, the purple C. violatinctus from Guam, as well as asci- 1987; Kott, 1990). Thus, all the potential defense didemin, the major alkaloid of the blue morph, signif- mechanisms are located in separate compartments. In icantly deter fish but not sea urchin predation. In addition, acidity by itself or in combination with spi- contrast, acidity and spicules by themselves deterred cules did not deter fish feeding, although fish were neither fish nor urchin feeding. All crude extracts gave observed to repeatedly regurgitate and swallow the similar results, although the alkaloid composition of the food before they finally consumed it. When present, purple forms (from Guam and the Mediterranean) and calcareous spicules may cause an extremely rapid neu- the blue Mediterranean morph were qualitatively dif- tralization of acid and, even without spicules, seawater ferent in their pyridoacridine composition (see Intro- may be a sufficient buffer to quickly neutralize the acid duction). Our results suggest, therefore, that both (Parry, 1984) and thus allow fish to feed. chemotypes have evolved into an effective defense Burns and Ilan (2003) found that deterrence in mechanism against predation. sponges was linked to spicule size, and only those Crude extracts alone significantly deterred damsel- spicules larger than 250 Am deterred predation. Chanas fish feeding in field assays and puffer fish feeding in the and Pawlik (1995) found no significant deterrence laboratory. However, while damselfish left many cubes caused by siliceous spicules in sponges and suggested nearly untouched, puffer fish repeatedly sampled the a deterrent role of sclerites of calcium carbonate related offered food. Puffer fish are benthic predators and to an alteration of pH in the acidic gut of putative occasionally feed on ascidians (Amesbury and Myers, predators. A similar defense mechanism was previously 1982). Therefore, they may be more used or tolerant to suggested by Schupp and Paul (1994) for calcified amino acid-derived metabolites such as alkaloids, the algae. The disc-shaped spicules, up to 1 mm in diam- main defensive compounds in ascidians (Davidson, eter, characteristic of the genus Cystodytes, represent 1993; Molinski, 1993). However, the addition of acid 0.64% of the wet mass of the ascidian. However, as to the combination of ascididemin and spicules made observed by Lindquist et al. (1992) for other ascidian 34 S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 species, spicular composition and concentration did not meaning that there is always scope for the evolution of significantly deter fish or sea urchin predation. In arti- specialist predators able to circumvent the defense ficial food, spicules were distributed at random, whilst mechanisms of a given species. in the living ascidian they encase each zooid. In fact, In conclusion, secondary metabolites seemed the spicules form a compact and relatively hard capsule that most effective defense mechanism of Cystodytes spp. lodges around the abdomen and into which the thorax against fish predators but did not deter sea urchins. can be retracted as a response to any disturbance. Although spicules and tunic acidity may perform Therefore, although the disc-shaped spicules did not other ecological roles and/or target certain specialist act as a deterrent in artificial food, they may help predators, our inability to keep low pH for more than protect the zooid in living colonies. In contrast, no 1 h and to arrange the spicules in capsules inside role could be attributed to the spherical spicules artificial food prevented us from drawing definite con- found in the Mediterranean purple morph, as they clusions as to their ecological role. Our results and measured only ~70 Am in diameter, are scarce and are other studies of sponges and gorgonians (Pawlik et randomly distributed throughout the tunic (Lo´pez- al., 1995; Koh et al., 2000; O’Neal and Pawlik, 2002; Legentil and Turon, 2005). Calcareous spicules may Puglisi et al., 2002) indicate that secondary metabolites also function as a shield to protect zooids from the are the primary means of defense against fish predators. acid release that may occur during cell rupture. This In addition, our study highlights the importance of suggestion was supported by stereomicroscope obser- considering all possible defenses of an organism against vation of bubbling in broken colonies, presumably as many potential predators as possible, in order to caused by a neutralization of acid as a result of calcium reliably assess their ecological roles. carbonate (spicules) dissolution. Previous studies found that chemical defenses de- Acknowledgements terred feeding by some, but not all, predators assayed (Paul and Hay, 1986; Hay et al., 1987; Schupp and Allison Palmer and Aja Reyes (Marine Lab, GU, Paul, 1994; Pisut and Pawlik, 2002; Tarjuelo et al., USA), and Raphael Ritson-Williams (Smithsonian Ma- 2002; Burns et al., 2003). Similarly, our results show rine Station, FL, USA) helped with the feeding assays. that all crude extracts of Cystodytes deter fish feeding, Dr. Bernard Banaigs and Gemma Agell made helpful but not sea urchin grazing. Indeed, urchin feeding comments on an early version of the article. This study seemed to be based on how likely they were to find was funded by the Education and Science Ministry the screen, rather than on any deterrence associated (MEC) and the project CTM2004-05265 of the Spanish with the crude extracts, ascididemin or the spicules. Government, NIH MBRS SCORE S06-GM-44796-14- Acidity is commonly found in combination with other 2 grant to P. Schupp, and the INTERREG IIIA no. I3A- defense mechanisms, and Stoecker (1980a) and Pelle- 1-72-E program of the EU. [RH] treau and Muller-Parker (2002) suggested that acidity had a major deterrent function against crawling preda- tors, because of the rupture of acid vesicles when they References contacted the ascidian surface. As we could not test acidity with sea urchins, a role for acidity in deterring Amesbury, S.S., Myers, R.F., 1982. The Fishes: I. Guide to the sea urchin predation cannot be ruled out. However, we Coastal Resources of Guam. University of Guam Press, Mangilao, pp. 1–141. have occasionally observed urchin bites in the purple Barnes, R.D., 1987. The Sponges. Invertebrate Zoology. CBS College and blue morphs of Cystodytes from the Mediterranean. Publishing, Philadelphia. Consumption by Mediterranean sea urchins was con- Becerro, M.A., Turon, X., Uriz, M.J., 1997. Multiple functions for firmed by the observation of disc-shaped spicules in the secondary metabolites in encrusting marine invertebrates. gut of urchins close to grazed Cystodytes spp. colonies. J. Chem. Ecol. 23, 1527–1547. Becerro, M.A., Paul, V.J., Starmer, J., 1998. Intracolonial variation in As well as deterring predators, defense mechanisms chemical defenses of the sponge Cacospongia sp. and its con- often act at other levels, such as fouling avoidance or sequences on generalist fish predators and the specialist nudi- space competition (Stoecker, 1980a; Schmitt et al., branch predator Glossodoris pallida. Mar. Ecol. Prog. Ser. 168, 1995; Becerro et al., 1997). Furthermore, they may 187–196. act at different life history stages, as suggested by Bontemps, N., 1996. Noyau pyridoacridine, structure et synthe`se d’alcaloı¨des cytotoxiques isole´s d’inve´rte´bres marins. PhD thesis. Uriz et al. (1996) and Pisut and Pawlik (2002). At the University of Perpignan, 1–160. same time, their multiple functions may constrain their Bowden, B.F., 2000. Aromatic alkaloids from ascidians. St. Nat. Prod. evolution (Schmitt et al., 1995; Kubanek et al., 2002), Chem. 23, 233–283. S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36 35

Bracher, F., 1989. Total synthesis of the pentacyclic alkaloid ascidi- Hay, M.E., Kappel, Q.E., Fenical, W., 1994. Synergisms in plant- demin. Heterocycles 29, 2093–2095. defenses against herbivores: interactions of chemistry, calcifica- Briscoe, C.S., Sebens, K.P., 1988. Omnivory in Stronglyocentrotus tion, and plant quality. Ecology 75, 1714–1726. droebachiensis (Mu¨ller) (Echinodermata: Echinoidea): predation Hill, M.S., Lopez, N.A., Young, K.A., 2005. Anti-predator defenses on subtidal mussels. J. Exp. Mar. Biol. Ecol. 115, 1–24. in western North Atlantic sponges with evidence of enhanced Burns, E., Ilan, M., 2003. Comparison of anti-predatory defenses of defense through interactions between spicules and chemicals. Red Sea and Caribbean sponges: II. Physical defense. Mar. Ecol. Mar. Ecol. Prog. Ser. 291, 93–102. Prog. Ser. 252, 115–123. Hirose, E., 1992. Tunic cells in Leptoclinides echinatus (Didemnidae, Burns, E., Ifrach, I., Carmeli, S., Pawlik, J.R., Ilan, M., 2003. Com- Ascidiacea): an application of scanning electron microscopy for parison of anti-predatory defenses of Red Sea and Caribbean paraffin embedding specimens. Hiyoshi. Rev. Nat. Sci. 11, 5–8. sponges: I. Chemical defense. Mar. Ecol. Prog. Ser. 252, 105–114. Kobayashi, J., Cheng, J., Nakamura, H., Ohizumi, Y., Hirata, Y., Caralt, S., Lo´pez-Legentil, S., Tarjuelo, I., Uriz, M.J., Turon, X., Sasaki, T., Ohta, T., Nozoe, S., 1988. Ascididemin, a novel 2002. Contrasting biological traits of Clavelina lepadiformis pentacyclic aromatic alkaloid with potent antileukemic activity (Ascidiacea) populations from inside and outside harbours in the from the Okinawan tunicate Didemnun sp. Tetrahedron Lett. 29, western Mediterranean. Mar. Ecol. Prog. Ser. 244, 125–137. 1177–1180. Caroll, A.R., Scheuer, P.J., 1990. Kuanoniamines A, B, C and D: Koehl, M.A.R., 1982. Mechanical design of spicule-reinforced con- pentacyclic alkaloids from a tunicate and its prosobranch mollusc nective tissue. J. Exp. Biol. 98, 239–268. predator Chelynotus semperi. J. Org. Chem. 55, 4426–4431. Koh, L.L., Goh, N.K.C., Chou, L.M., Tan, Y.W., 2000. Chemical and Carroll, A.R., Cooray, N.M., Poiner, A., Scheuer, P.J., 1989. A second physical defenses of Singapore gorgonians (Octocorallia: Gorgo- Shermilamine alkaloid from a tunicate Trididemnum sp. J. Org. nacea). J. Exp. Mar. Biol. Ecol. 251, 103–115. Chem. 54, 4231–4232. Kott, P., 1990. The Australian ascidiacea: Part 2. Aplousobranchia (1). Chanas, B., Pawlik, J.R., 1995. Defenses of Caribbean sponges Mem. Queensl. Mus. 29, 1–266. against predatory reef fish: II. Spicules, tissue toughness, and Kubanek, J., Whalen, K.E., Engel, S., Kelly, S.R., Henkel, T.P., nutritional quality. Mar. Ecol. Prog. Ser. 127, 195–211. Fenical, W., Pawlik, J.R., 2002. Multiple defensive roles for Dassonneville, L., Wattez, N., Baldeyrou, B., Mahieu, C., Lansiaux, triterpene glycosides from two Caribbean sponges. Oecologia A., Banaigs, B., Bonnard, I., Bailly, C., 2000. Inhibition of 131 (1), 125–136. topoisomerase II by the marine alkaloid ascididemin and induc- Lambert, G., 1980. Predation by the prosobranch mollusk Lamellaria tion of apoptosis in leukemia cells. Biochem. Pharmacol. 60, diegoensis on Cystodytes lobatus, a colonial ascidian. Veliger 22, 527–537. 340–344. Davidson, B.S., 1993. Ascidians: producers of amino acid derived Lambert, G., 2003. Marine biodiversity of Guam: the Ascidiacea. metabolites. Chem. Rev. 93, 1771–1791. Micronesica 35–36, 588–597. Davis, A.R., 1991. Alkaloids and ascidian chemical defense: evidence Lewis, J.C., Von Wallis, E., 1991. The function of surface scler- for the ecological role of natural products from Eudistoma oliva- ites in gorgonians (Coelenterata, Octocorallia). Biol. Bull. 181, ceum. Mar. Biol. 111, 375–379. 275–288. Davis, A.R., Wright, A.E., 1989. Interspecific differences in fouling Lindquist, N., Hay, M.E., Fenical, W., 1992. Defense of ascidians and of two congeneric ascidians (Eudistoma olivaceum and E. capsu- their conspicuous larvae: adults vs larval chemical defenses. Ecol. latum): is surface acidity an effective defense? Mar. Biol. 102, Monogr. 62, 547–568. 491–497. Lindsay, B.S., Barrows, L.R., Copp, B.R., 1995. Structural require- Debard, H., Banaigs, B., Francisco, F., Commeyras, A., 1998. Use of ments for biological activity of the marine alkaloid ascididemin. ascididemin and derivatives as antifouling agents. PCT Int. Appl. Bioorg. Med. Chem. Lett. 5, 739–742. WO 98/21959. Lo´pez-Legentil, S., 2005. Multidisciplinary studies of the genus Della Valle, A., 1877. Contribuzioni alla stri naturale delle Ascidie Cystodytes (Ascidiacea): from molecules to species. PhD thesis, Composte del Golfo di Napoli. Napoli. University of Barcelona and Perpignan, 1–217. Eder, C., Schupp, P., Proksch, P., Wray, V., Steube, K., Mu¨ller, C.E., Lo´pez-Legentil, S., Turon, X., 2005. How do morphoytpes and Frobenius, W., Herderich, M., van Soest, R.W.M., 1998. Bioactive chemotypes relate to genotypes? The colonial ascidian Cystodytes pyridoacridine alkaloids from the micronesian sponge Oceanapia (Ascidiacea: Polycitoridae). Zool. Scr. 34 (1), 3–14. sp. J. Nat. Prod. 61, 301–305. Lo´pez-Legentil, S., Dieckmann, R., Bontemps-Subielos, N., Turon, Fretter, V., Graham, A., 1962. British Prosobranch Molluscs, Their X., Banaigs, B., 2005a. Qualitative variation of alkaloids in color Functional Anatomy and Ecology. Ray Society, London, England. morphs of Cystodytes (Ascidiacea). Biochem. Syst. Ecol. 33 (11), Goodbody, I., Gibson, J., 1974. The Biology of Ascidia nigra 1107–1119. (Savigny): V. Survival in populations settled at different times Lo´pez-Legentil, S., Ruchy, M., Dome´nech, A., Turon, X., 2005b. Life of the year. Biol. Bull. 146, 217–237. cycles and growth rates of two morphotypes of Cystodytes (Asci- Harvell, C.D., Fenical, W., 1989. Chemical and structural defenses of diacea) in the western Mediterranean. Mar. Ecol. Prog. Ser. 296, Caribbean gorgonians (Pseudopterogorgia spp.): intracolony lo- 219–228. calization of defense. Limnol. Oceanogr. 34, 382–389. McClintock, J.B., Heine, J., Slattery, M., Weston, J., 1991. Biochem- Harvell, C.D., Fenical, W., Greene, D.H., 1988. Chemical and struc- ical and energetic composition, population biology, and chemical tural defenses of Caribbean gorgonians (Pseudopterogorgia spp.): defense of the Antarctic ascidian Cnemidocarpa verrucosa Les- I. Development of an in situ feeding assay. Mar. Ecol. Prog. Ser. son. J. Exp. Mar. Biol. Ecol. 147, 163–175. 49, 287–294. Millar, R.H., 1971. The biology of ascidians. Adv. Mar. Biol. 9, Hay, M.E., Duffy, J.E., Pfister, C.A., Fenical, W., 1987. Chemical 1–100. defenses against different marine herbivores: are amphipods insect Molinski, T.F., 1993. Marine pyridoacridine alkaloids: structure, syn- equivalents? Ecology 68, 1567–1580. thesis, and biological chemistry. Chem. Rev. 93, 1825–1838. 36 S. Lo´pez-Legentil et al. / Journal of Experimental Marine Biology and Ecology 332 (2006) 27–36

Monniot, F., 1988. Ascidies de Nouvelle-Cale´donie: V. Polycitoridae Schmitt, T.M., Hay, M.E., Lindquist, N., 1995. Constraints on chem- du lagon. Bull. Mus. Hist. Nat. 4 (10), 197–235. ically mediated coevolution: multiple functions for seaweed sec- Morris, R.H., Abbott, D.P., Haderlie, E.C., 1980. Intertidal Inverte- ondary metabolites. Ecology 76 (1), 107–123. brates of California. Stanford University Press, Stanford, Schmitz, F.J., DeGuzman, F.S., Hossain, M.B., van der Helm, D., pp. 177–226. 1991. Cytotoxic aromatic alkaloids from the ascidian Amphicarpa Myers, R.F., 1983. The comparative ecology of the shallow-water meridiana and Leptoclinides sp. Meridine and 11-Hydroxyasci- species of Canthigaster (Family ) of Guam. MSc didemin. J. Org. Chem. 56, 804–808. thesis, University of Guam. Schupp, P., Paul, V.J., 1994. Calcium carbonate and secondary meta- O’Neal, W., Pawlik, J.R., 2002. A reappraisal of the chemical and bolites in tropical seaweeds: variable effects on herbivorous fishes. physical defenses of Caribbean gorgonian corals against predatory Ecology 75, 1172–1185. fishes. Mar. Ecol. Prog. Ser. 240, 117–126. Schupp, P., Wray, V., Eder, C., Schneider, P., Herderich, M., Paul, Parry, D.L., 1984. Chemical properties of the test of ascidians in V.J., Proksch, P., 1999. Staurosporine derivatives from the ascid- relation to predation. Mar. Ecol. Prog. Ser. 17, 279–282. ian Eudistoma toealensis and its predatory Pseudoceros Paul, V.J., Hay, M.E., 1986. Seaweed susceptibility to herbivory: sp. J. Nat. Prod. 62, 959–962. chemical and morphological correlates. Mar. Ecol. Prog. Ser. 33, Stoecker, D., 1978. Resistance of a tunicate to fouling. Biol. Bull. 255–264. 155, 615–626. Paul, V.J., Van Alstyne, K.L., 1992. Activation of chemical defenses Stoecker, D., 1980a. Chemical defenses of ascidians against preda- in the tropical green algae Halimeda spp. J. Exp. Mar. Biol. Ecol. tors. Ecology 61, 1327–1334. 160, 191–203. Stoecker, D., 1980b. Relationships between chemical defense and Paul, V.J., Lindquist, N., Fenical, W., 1990. Chemical defense of the ecology in benthic ascidians. Mar. Ecol. Prog. Ser. 3, 257–265. tropical ascidian Atapazoa sp. and its nudibranch predators Nem- Stoecker, D., 1980c. Distribution of acid and vanadium in Rhopalaea brotha sp. Mar. Ecol. Prog. Ser. 17, 279–282. birkelandi Tokioka. J. Exp. Mar. Biol. Ecol. 48, 277–281. Pawlik, J.R., Fenical, W., 1992. Chemical defense of Pterogorgia Swinehart, J.H., Biggs, W.R., Halko, D.J., Schroder, N.C., 1974. The anceps, a Caribbean gorgonian coral. Mar. Ecol. Prog. Ser. 97, vanadium and selected metal contents of some ascidians. Biol. 183–188. Bull. 146, 302–312. Pawlik, J.R., Burch, M.T., Fenical, W., 1987. Patterns of chemical Tarjuelo, I., Lo´pez-Legentil, S., Codina, M., Turon, X., 2002. Defence defense among Caribbean gorgonian corals: a preliminary survey. mechanisms of adults and larvae of marine invertebrates: patterns J. Exp. Mar. Biol. Ecol. 108, 55–66. of toxicity and palatability in colonial ascidians. Mar. Ecol. Prog. Pawlik, J.R., Chanas, B., Toonen, R.J., Fenical, W., 1995. Defense of Ser. 235, 103–115. Caribbean sponges against predatory reef fish: I. Chemical deter- Turon, X., 1987. Estudio de las ascidias de las costas de Catalun˜a e rency. Mar. Ecol. Prog. Ser. 17, 183–194. Islas Baleares. PhD thesis, University of Barcelona, 1–353. Pelletreau, K.N., Muller-Parker, G., 2002. Sulfuric acid in the phaeo- Turon, X., Lo´pez-Legentil, S., Banaigs, B., 2005. Cell types, micro- phyte alga Desmarestia munda deters feeding by the sea urchin symbionts and pyridoacridine distribution in the tunic of Cysto- Strongylocentrotus droebachiensis. Mar. Biol. 141, 1–9. dytes (Ascidiacea, Polycitoridae). Inv. Biol. 124 (4), 355–369. Pennings, S.C., Paul, V.J., 1992. Effect of plant toughness, calcifica- Uriz, M.J., Turon, X., Becerro, M.A., Galera, J., 1996. Feeding tion, and chemistry on herbivory by Dolabella auricularia. Ecol- deterrence in sponges. The role of toxicity, physical defenses, ogy 73, 1606–1619. energetic contents, and life-history stage. J. Exp. Mar. Biol. Pisut, D.P., Pawlik, J.R., 2002. Anti-predatory chemical defenses of Ecol. 205, 187–204. ascidians: secondary metabolites or inorganic acids? J. Exp. Mar. Van Alstyne, K.L., Wylie, C.R., Paul, V.J., 1994. Antipredator Biol. Ecol. 270, 203–214. defenses in tropical Pacific soft corals (Coelenterata: Alcyonacea): Puglisi, M.P., Paul, V.J., Slattery, M., 2000. Biogeographic differences II. The relative importance of chemical and structural defenses in in physical and chemical defenses of the pacific gorgonians three species of Sinularia. J. Exp. Mar. Biol. Ecol. 178, 17–34. Annella mollis and A. reticulata. Mar. Ecol. Prog. Ser. 207, Vervoort, H.C., Pawlik, J.R., Fenical, W., 1998. Chemical defense of 263–272. the Caribbean ascidian Didemnum conchyliatum. Mar. Ecol. Prog. Puglisi, M.P., Paul, V.J., Biggs, J., Slattery, M., 2002. Co-occurrence Ser. 164, 221–228. of chemical and structural defenses in gorgonian corals from Webb, D.A., 1939. Observations on the blood of certain ascidians, Guam. Mar. Ecol. Prog. Ser. 239, 105–114. with special reference to the biochemistry of vanadium. J. Exp. Randall, J.E., Hartman, W.D., 1968. Sponge feeding fishes of the Biol. 16, 499–523. West Indies. Mar. Biol. 1, 216–225. Wylie, C.R., Paul, V.J., 1989. Chemical defenses in three species of Rogers, S.D., Paul, V.J., 1991. Chemical defenses of three Glosso- Sinularia (Coelenterata, Alcyonacea): effects against generalist doris nudibranchs and their dietary Hyrtios sponges. Mar. Ecol. predators and the butterflyfish Chaetodon unimaculatus Bloch. Prog. Ser. 77, 221–232. J. Exp. Mar. Biol. Ecol. 129, 141–160.